Serious Doubts in Renormalization in HFSS

[mirriyaz]

deembedding in hfss

Hiiiiiiiiii Guys,

I have a doubt regarding the Renormalization and De-Embed available in waveport assignment dialogue box of HFSS, my doubt is what is exactly Renormalization and De-Embedding, and why or when do we use it.

I have some knowledge and I have used it, like I designed a single p@tch microstrip antenna for 10GHz and am feeding it with microstripline of 50 ohms, so I set the renormalization for the wave port to 50 ohms and de-embed to 0, but after the simulation my response is not at 10GHz but different, so when I change the 50 ohms in post processing it gives me different response which is again not at 10GHZ, so I'm kind of stuck, if anyone can plzz help me as to when you have to renormalize and how much you have to De-Embed, it'll be really helpful, as far as I know my calculations for p@tch dimensions are correct.

thanx in advance.

 

[borich_03]

hfss post processing renormalize

If you use a WavePort, may be you need try to reassign the Port. Use different Port's configuration, Terminal Lines. You can use Driven Terminal Method. And at the first, try to simulate "Only port solution".
Also, test about higher-order modes. Ansoft recommends: "..Wave Port Require a Lenght of Uniform Cross Section. For example: Each port you define connected to a semi-infinitely long waveguide that has the same cross section as the Wave Port."

 

[loucy]

"as far as I know my calculations for p@tch dimensions are correct"=>did you use some simple formula to calculate the dimension for the patch so that it "should/would" resonate at 10GHz? If so, then cosider that there are various assumptions involved with the formula. for example, the loading effect of the feed is not accounted for; fringing effect at the edge is not accurately predicted, etc. As a result, the resonant frequency predicted by the simple formula is not good. You can't use it as a "standard/benchmark" for judging the accuracy of HFSS simulation in general. In particular, the effect of de-embedding is seldom considered in a simplified model.

for your problem, renormalization is kind of like attaching a source with different internal resistance to the antenna. naturally you would expect to see different level of radiation, i.e. the location and minimum of |S11| would be different.

my suggestion: use wave port, do not apply renormalization, look at input impedance (rather than S11) to find the resonant frequency.

 

[akhanm]

i have some problems about waveport of rectangular microstrip antenna..